The AI Divide: Why 70% of Media Campaigns Haven’t Evolved (And How the Other 30% Are Winning)

Picture this: You’re sitting in a campaign planning meeting, and someone suggests using AI to build your media strategy from scratch. Five years ago, that person would’ve been politely ignored. Today? They might just be onto something big.

The advertising world is having a moment with AI, but it’s not the smooth sailing story you might expect. According to the new IAB “State of Data 2025: The Now, The Near, and The Next Evolution of AI for Media Campaigns“ report, a surprising 70% of agencies, brands, and publishers still haven’t fully integrated AI across their media planning, activation, and analysis processes. That’s right, despite all the buzz, most companies are still figuring this out.

But here’s the plot twist: those who are getting it right aren’t just saving time and money. They’re fundamentally changing how campaigns work.

The New AI Playbook: Beyond Just “Smarter Ads”

Remember when AI in advertising meant slightly better targeting and maybe some automated bidding? Those days are over. Today’s AI doesn’t just optimize existing processes, it thinks, strategizes, and executes in ways that would make your 2019 media planner’s head spin.

We’re talking about AI that can:

• Build complete media plans from scratch
• Generate audience segments using synthetic data
• Select media partners and create customized sales packages
• Run scenario planning and forecast performance in real-time
• Handle cross-channel bid optimization while you sleep

One brand supervisor put it perfectly:

AI is great when you need to make real-time, specific, high-volume decisions. It does not do a good job at providing larger scale, strategic guidance, or recommendations that don’t skew towards what’s already working.

That balance between tactical brilliance and strategic limitation is exactly where the industry finds itself right now.

The Adoption Race: Who’s Winning and Who’s Lagging

Here’s where it gets interesting. Agencies and publishers are absolutely crushing it with AI adoption, while brands are… well, they’re trying their best.

Agencies are leading the charge, with 51% already using AI for audience identification and segmentation. They’re also getting creative with generative AI, using it to build segments with synthetic data, basically creating fake but statistically accurate audience profiles to test strategies without touching real customer data.

Publishers are focusing on the crystal ball stuff, 41% are using AI for inventory prediction and forecasting. They’re also getting smart about attribution, tracking engagement across devices and platforms in ways that would’ve required small armies of analysts just a few years ago.

Brands, meanwhile, are still mostly in pilot mode. They’re being cautious, and honestly, for good reason. When you’re the one writing the checks and explaining ROI to the C-suite, “let’s see what happens” isn’t always a viable strategy.

The Real Roadblocks (Hint: It’s Not About Jobs)

You might expect the biggest concern to be “Will AI replace our jobs?” Surprisingly, that’s dead last on the worry list, mentioned by only 37% of respondents. The real headaches are much more practical:

Data chaos tops the list. Companies are drowning in data quality issues, security risks, and the nightmare of trying to get different AI tools to talk to each other. It’s like trying to conduct an orchestra where half the musicians are playing different songs.

The complexity monster is close behind. Setting up and maintaining AI systems isn’t exactly plug-and-play. We’re talking about 62% of companies struggling with implementation complexity.

The knowledge gap is real too. Teams know AI could help, but they’re not always sure how to make it happen effectively.

What’s particularly fascinating is the transparency tension brewing between brands and agencies. Half of brands worry they don’t really understand how their partners are using AI on their behalf. Meanwhile, half of agencies are nervous that brands will just build AI capabilities in-house and cut them out of the equation.

Talk about trust issues.

What the AI Winners Are Actually Doing

Despite all these challenges, some companies are making it work. And they’re not just throwing technology at problems and hoping for the best.

The smart ones are building phased roadmaps. They’re mapping out AI use cases across their entire campaign lifecycle, setting clear milestones, and, this is crucial, actually updating their plans as AI capabilities evolve. Because let’s be honest, what AI could do six months ago looks quaint compared to today.

Data readiness is their obsession. They’re auditing data quality, beefing up security, and requiring human oversight to catch AI when it gets creative in unhelpful ways. They’re also using synthetic data for testing, essentially creating safe sandbox environments for AI experimentation.

Training is huge. The companies getting real value from AI aren’t just buying software; they’re investing in their people. Structured training programs, shared resources like prompt libraries, and cross-functional AI centers of excellence are becoming standard practice.

The Numbers That Tell the Real Story

Here’s what effectiveness looks like in practice:

• 82% of companies say AI meets expectations for efficiency
• 75% report it’s effective at what it’s supposed to do
• 73% find it reliable enough to depend on

But AI isn’t equally good at everything. It absolutely dominates precision tasks, audience segmentation and data aggregation are where it shines brightest. But ask it to prevent ad fraud, manage brand safety, or handle contract negotiations? You’re better off sticking with humans for now.

The 2026 Tipping Point

Here’s the timeline that should get your attention: half of the companies not yet using AI at scale expect to have full integration by 2026. That’s not that far away.

This means we’re looking at a potential inflection point where AI moves from “nice to have” to “competitive necessity” in just a couple of years. The companies that figure this out first aren’t just going to save money, they’re going to have fundamentally different capabilities than their competitors.

What This Means for Your Next Campaign

If you’re running media campaigns and feeling overwhelmed by AI possibilities, you’re not alone. But you also can’t afford to wait until everyone else figures it out.

Start with the basics: audit your data quality, identify one or two specific use cases where AI could make an immediate difference, and begin building internal knowledge. Whether that’s using AI for audience segmentation, real-time bid optimization, or performance forecasting, pick something concrete and learn from there.

Most importantly, focus on building trust and transparency with your partners. Whether you’re a brand working with agencies or an agency managing multiple clients, clear communication about AI capabilities and limitations will matter more than the technology itself.

The companies winning with AI today aren’t the ones with the fanciest tools, they’re the ones who’ve figured out how to make AI and human expertise work together effectively. And that’s a challenge worth solving, whether you’re in that 70% still figuring it out or part of the group already seeing results.

The AI revolution in media campaigns isn’t coming, it’s here. The only question is whether you’ll be leading it or playing catch-up.

Frequently Asked Questions

Q: What’s actually new about AI in media campaigns? Haven’t we been using it for years?

A: While AI has been around for data analysis and basic optimization, we’re now seeing generative and agentic AI that can actually think strategically. Instead of just helping with existing processes, AI can now build complete media plans, generate audience segments, select partners, and even create synthetic data for testing, essentially handling tasks that previously required human strategists.

Q: Who’s really ahead in AI adoption right now?

A: Agencies and publishers are leading the charge, driven primarily by efficiency gains and cost savings. They’re more likely to have organization-wide AI licenses and custom solutions. Brands are typically still in pilot phases, largely due to resource constraints and the pressure to prove ROI before scaling up.

Q: What’s the biggest barrier to AI adoption, is it really about job security?

A: Surprisingly, no. Job displacement ranks lowest among concerns at just 37%. The real challenges are data readiness (quality, security, accessibility), system complexity, tool fragmentation, and lack of AI knowledge. Most companies are more worried about getting AI to work properly than about it working too well.

Q: Where does AI actually work well versus where it falls short?

A: AI excels at precision tasks and data consolidation, think audience segmentation, data aggregation, real-time optimizations, and targeting. It struggles where human judgment and strategic thinking are crucial, like ad fraud prevention, brand safety monitoring, contract management, and providing strategic guidance that goes beyond established patterns.

Q: What should companies focus on first for successful AI adoption?

A: Start with a phased roadmap that maps your specific AI use cases. Ensure your data is clean, secure, and properly governed. Invest heavily in training your team, AI tools are only as good as the people using them. Most importantly, pick one or two specific use cases to master before trying to transform everything at once.